search

PWhiddy / PokemonRedExperiments

Playing Pokemon Red with Reinforcement Learning
MIT License
6.99k stars 644 forks source link

Where do the models originate? (question) #149

Closed ethicalhacker7192 closed 11 months ago

ethicalhacker7192 commented 11 months ago

Hi, I was looking and trying out your code and I was wondering where your models came from, I am very curious how you got pyboy working together with gym as well.

PWhiddy commented 11 months ago

Hi! The model in the base version is the default CNN from stablebaselines (NatureCNN). As for how pyboy and gym are working together, see this file https://github.com/PWhiddy/PokemonRedExperiments/blob/master/baselines/red_gym_env.py

ethicalhacker7192 commented 11 months ago

just as a tip in github instead of doing a new line and doing your link, do [yoursubject] (yourlink) of course without the space.